[ANN] Smalltalk Chronicles

I am pleased to announce the return of Smalltalk Chronicles at its new URL
as http://www.smalltalkchronicles.net.

Comments, offers of help, and most especially articles, are welcomed.

--
Alan Knight [|], Cincom Smalltalk Development
[hidden email]
[hidden email]
http://www.cincom.com/scripts/smalltalk.exe/downloads/index.asp

 From a programmer's perspective, I can relate well with the
issues and arguments of the following article. Well done.

http://www.smalltalkchronicles.net/edition3-1/whyjava.html

However, from a manager's perspective and if I have no or even
a little programming background. The "bad movie" analogy doesn't
work. While most people can identify what a bad movie is. I am
not convinced the majority of project manangers can come to the
the same conclusion about programming languages.

I would think the article is making a case FOR Java because the
manager can produce a bigger budget (more $$$) for the project.
Doesn't all manager want a bigger budget? Why would I want
to budget a project to complete in 6 months when I can budget
for 18 months? I can create more work for my staff and give them
jobs for a longer duration.

Please excuse my sarcasm and I hope I am wrong.

Regards,

Long On
Edgesoft Consulting Inc.


"Alan Knight" <[hidden email]> wrote in message
news:9075860CFknightacmorg@24.2.9.58...

Heck, I'm a programmer and *I* can't relate to the analogy.  I'm iffy
on what the author calls "aesthetics", and I'm not sure what it has to
do with the paper.  Happily, though, the analogy is mostly ignored.

The meat of the paper is an excellent argument for why *simplicity* in
a language is important.  Even if you don't read the whole thing, it's
worth looking at the Smalltalk vs. Java decision graphs about 2/3
down.  Also, there is a nice manager-friendly graph of expertise over
time, paralleling the argument that most programmers aren't at their
optimum and thus that faster learning curves are important.


>
> I would think the article is making a case FOR Java because the
> manager can produce a bigger budget (more $$$) for the project.
> Doesn't all manager want a bigger budget? Why would I want
> to budget a project to complete in 6 months when I can budget
> for 18 months? I can create more work for my staff and give them
> jobs for a longer duration.
>



heheh!


-Lex

+---- [hidden email] wrote (Mon, 02 Apr 2001 15:47:02 GMT):
 | From a programmer's perspective, I can relate well with the
 | issues and arguments of the following article. Well done.
 |
 | http://www.smalltalkchronicles.net/edition3-1/whyjava.html
 +----

I disagree.  It's the same whiny, unbalanced, trivia based rant
against Sun Java that has failed many times in the past.
Smalltalk is not failing because of the success of Sun Java;
the sooner that y'all grow up the better off Smalltalk will be . . .

--
Gary Johnson     [hidden email]
Privacy on the net is still illegal.
<a href=http://www.squeak.org>Tired of selfish technology monopolies?</a>

Exactly, your "reality" and mine are definitely not the same.

Having lived and breathed Smalltalk for a number of years
now (mind you with other procedural languages as well), I
would still prefer to live within the Smalltalk VM over a JVM.

I have also been keeping up with Java development and recently
attempted a project using the latest Java release. After a week of
struggling to work around the constraints and short comings of Java,
I switched back to my favourite VM and was able to accomplish most
of what I have wanted the following week.

The whyjava article just confirms my belief that Java is still
not ready for prime-time and it does not look like it will be
for any time soon.

Please run a diag on your reality...are you sure it is the one you want?

Long On
Edgesoft Consulting Inc.


"Reality is a point of view" <[hidden email]> wrote in message
news:[hidden email]...

[hidden email] (Reality is a point of view) writes:


Whiny, yes.  Unbalanced, yes.  Trivial?  No way!  Managers tend to
look at one languages being equal to another, and at any task being
doable in any language.  From a distant enough perspective, this is
true, but this article makes an excellent argument that the complexity
of using different languages has observably large effects, even if
it's a manager doing the observing.

It's not "this language is pure" or "this language is elegant".  It
does say "this language is aesthetic", which I agree is silly, but the
rest of the article is about complexity and learning curves.

What do you think of the decision trees?  Do they look realistic?  And
what about the argument about programmers never reaching full
expertice?

-Lex

"Lex Spoon" <[hidden email]> wrote in message

> It's not "this language is pure" or "this language is elegant".  It
> does say "this language is aesthetic", which I agree is silly,

I can certainly relate to the bad movie analogy. Maybe there are good
reasons to say that it is "silly". I don't care. There are differences. With
the right frame of mind, one can greatly enjoy a bad movie, though ;-).

> but the
> rest of the article is about complexity and learning curves.

That is very interesting, even though it's mostly theory.

> What do you think of the decision trees?  Do they look realistic?

Yes, they are not only realistic, but also insightful. They are the best
part of the article. To me they are extra interesting, because they also
make clear what improvements (i.e., simpler decision trees) one can still
look for relative to Smalltalk (which is one of my favorite research areas).

> And
> what about the argument about programmers never reaching full
> expertice?

A very important point, and well made.

Regards,

Peter van Rooijen

Reality is the last refuge of the unimaginative :)

I didn't find the article whiny or unbalanced, but those are certainly
subjective rankings.

Smalltalk gets the job done, Java may get the job done -- I don't have
direct experience -- but I agree with the argument that Smalltalk is
technically superior to Java.

The thing we all have to remember is that technical excellence has *NOTHING*
to do with market success.

in article [hidden email], Reality is a point of
view at [hidden email] wrote on 4/2/01 12:54 PM:

I have plenty of experience with both languages and I wholeheartedly
agree with Sigrid (the author). And yet I end up working in Java
nowadays because it does have its own "presence" in the IS shops
throughout the lands.

For those that haven't had the .pleasure. of using both languages, I
would suggest you try that in order to balance your own perspective on
the matter. But then again, I have played around with plenty of other
interesting languages - like Eiffel, Prograph, Perl, Python, Ruby, etc.


And even though I really like Smalltalk, I know that Alan Kay's
inspiration for the language was somewhat different than what Smalltalk
has ended up being used for (at least outside of Disney). He wanted to
make the DynaBook happen, which as far as I can tell, we're still
waiting for.

The folks at PARC were ahead of the rest of the industry back then, and
even now. It's not too hard when the best you can expect is an OS called
ME! And languages with mucho keywords, etc.



In article <B6EEA45B.E2D9%t*b*r*u*m*l*e*y@f*u*s*e.n*e*t>,
 Troy Brumley <t*b*r*u*m*l*e*y@f*u*s*e.n*e*t> wrote:

> Reality is the last refuge of the unimaginative :)
>
> I didn't find the article whiny or unbalanced, but those are certainly
> subjective rankings.
>
> Smalltalk gets the job done, Java may get the job done -- I don't have
> direct experience -- but I agree with the argument that Smalltalk is
> technically superior to Java.
>

Note: followup set.

 +---- [hidden email] wrote (02 Apr 2001 13:58:48 -0400):
 | [hidden email] (Reality is a point of view) writes:
 | >  +---- [hidden email] wrote (Mon, 02 Apr 2001 15:47:02 GMT):
 | >  | http://www.smalltalkchronicles.net/edition3-1/whyjava.html
 | >  +----
 | >
 | > I disagree.  It's the same whiny, unbalanced, trivia based rant
 | > against Sun Java that has failed many times in the past.
 | > Smalltalk is not failing because of the success of Sun Java;
 | > the sooner that y'all grow up the better off Smalltalk will be . . .
 |
 | Whiny, yes.  Unbalanced, yes.  Trivial?  No way!  Managers tend to

Trivia, not trivial.

 | true, but this article makes an excellent argument that the complexity
 | of using different languages has observably large effects, even if
 | it's a manager doing the observing.

I stopped reading the pseudo science when the apologist stance
became clear.

 | It's not "this language is pure" or "this language is elegant".  It
 | does say "this language is aesthetic", which I agree is silly, but the
 | rest of the article is about complexity and learning curves.

And it completely ignores those of Smalltalk as well as the
broader issue of Smalltalks continued, nay repeated, failure.

 | What do you think of the decision trees?  Do they look realistic?  And
 | what about the argument about programmers never reaching full
 | expertice?
 +----

I think the baby and the bath water need separating.  Focusing
on the 5 easy pieces at the expense of science is fraudulent.
How much of the hierarchy fits into the little five?  How much
of the Smalltalk variant trivia fits into the little five?  Was
their any discussion of the confusing operator precedence?

Recently someone poked holes in the pair programming 'science',
with predictable reactions from the community.

Until y'all accept the fact that Smalltalk has lost, and
continues to lose, and work to solve the outstanding problems
instead of making VHS/Beta excuses the better off Smalltalk,
and computer science, will be.

--
Gary Johnson     [hidden email]
Privacy on the net is still illegal.
<a href=http://www.squeak.org>Tired of selfish technology monopolies?</a>

[hidden email] says...
> Smalltalk is not failing because of the success of Sun Java;
> the sooner that y'all grow up the better off Smalltalk will be . . .

But it feels *so* much better to blame something else ;-)

Seriously though: I've read many of your postings and I try to understand
your point without letting their inflammatory nature cloud my judgment
(as both Alan Kay and Dave Thomas said: we are a smug group).

What does the Smalltalk community need to do to 'grow up'?  

I have a vested interest in making Smalltalk more popular, which is why I
keep the Toronto Smalltalk User Group going.  But Smalltalk's lack of
penetration into the larger development world is an ongoing source of
frustration.

We all seem to feel that Smalltalk is not a popular as it should be, and
that anyone that does an unbiased review would agree (smug smug smug).

What *should* we do to sell Smalltalk?
--
Bob Nemec
Newcastle Objects
[hidden email]

Bob Nemec <[hidden email]> wrote in
<[hidden email]>:

It needs to be promoted on several fronts.

One reason that Smalltalk is not easily accepted is that it is
not familiar.  To change this, we need to get the schools to teach
with it.  So it seems that a logical first step would be to
promote it with instructors and professors.  One way to do this
is to write more articles in journals that explain the features
of smalltalk and present its advantages.

--
Terry
===========================================================
Terry Raymond       Smalltalk Professional Debug Package
Crafted Smalltalk   *Breakpoints* and *Watchpoints* for
19 Tilley Ave.                  VW and ENVY/Developer
Newport, RI  02840
(401) 846-6573      [hidden email]
http://www.craftedsmalltalk.com
===========================================================

Terry Raymond wrote:
Another alternative is for those who feel they have the experience to go
out to their local community colleges and give a course or two in learning
or using Smalltalk- apply your knowledge of solving complex business problems
by illustrating at a high level how to do that in Smalltalk, or introduce
people to objects using Smalltalk.

I believe that this sort of activity could be organized as a sort of ongoing
CampSmalltalk project- Roger Whitney has had some good success in using Squeak
in his classes at San Diego State University, for instance.  It's not something
that would have to be done without some assistance from the Smalltalk advocacy
community.

Stop talking, start teaching!

- les

Terry Raymond <[hidden email]> wrote in message
news:Xns90857B042traymondcraftedsmall@207.126.101.100...
I couldn't agree more about the need to get Smalltalk into the schools
Terry.  There is however a nasty Catch 22 situation where schools will often
teach only that which they deem will be useful after graduation.  It also
seems that there are few instructors to teach the courses properly.  I think
you have a very good idea about putting out more articles.  There is an
impressive list of companies and projects that use Smalltalk.  People always
want to know who else is using something.  The shop I am with now had never
heard of Smalltalk before I came by.  They did alot of homework, and made a
decision regarding the development environment.

Earlier this year, I was talking to a major utility company that currently
uses Smalltalk in such an atrocious manner that projects are failing and
they are contemplating a switch to Java.  The sad part is that unless they
change development practices, the Java projects will also fail.  For now,
they are blaming Smalltalk.  There was a similar situation with a utility in
BC, Canada.  A really good idea gets a bad implementation and Smalltalk gets
a black eye.

John Gale
Procedium Software Corp.

What is really preventing wider acceptance of Smalltalk is exactly the same
thing that makes it so appreciated by the people who know it. Smalltalk in
all forms that I know it is a big, powerful system. Big, powerful systems
don't generally get widely used.

I'm not at all convinced that Smalltalk needs to be more widely used, but if
that's what you want, make it easy to program Smalltalk in Notepad, or
MS-Word.

If you know of anyone working on that, I'd sure be interested to hear about
it.

Regards,

Peter van Rooijen

Bob Nemec <[hidden email]> writes:
>
>What *should* we do to sell Smalltalk?

Its dead simple.
Smalltalk needs to be fast, small, inexpensive to get started with (even to
the stage of producing commercial products), cross platform, and have support
for common protocols and APIs (such as ODBC).

There are variants of Smalltalk that satisfy some or most of these
requirements, but none that meets them all.

For example:
Dolphin is inexpensive for commercial projects, but is not cross platform.
Squeak is cross platform, but of limited use commercially, and is slow.
VisualWorks is fast and cross platform, and great in terms of API support, but
expensive and has licencing issues for commercial work.

Steve Zara
>--
>Bob Nemec
>Newcastle Objects
>[hidden email]

When we started our first Smalltalk project I was amazed at the cost
of the software compared to Powerbuilder.  Parcplace had just stopped
charging runtime cost for the base stuff but we had to deal with the
runtime charges for DST.  Now standing on the Java side, Smalltalk is
cheap compared to what you are doing in JAVA.  A run down of the cost
to give you an example.  
Price out an EJB server per processor
        $4,000 to $15,000 per processor
If you go the Corba route
        $2500 to $15000 per processor
TOPLink for JAVA
        $7000 per processor
        (Alan ought to get a kick out of this.  Started using the
product last year repeatedly was told no runtime charges, Go to deploy
the app and find out they now implement runtime charges.  Webgain
claims they are not BEA but their licensing scheme has BEA written all
over it.  Know why Alan left!)

If you are using MSSQL throw in another couple of thousand for each
database so you can get a decent driver.  
Now look at developer charges
IDE
 $1000 to $4000 a developer
Toplink
 $3,500
ORB
 $3000
EJB
 $2,000 to $4,000

When Cincom finishes work on 5.4i and it includes the Smalltalk Server
pages, I will be able to do everything with one toolset that I
currently have to do with 4 to 5.  
Collections suck so you might as well get the JGL collections from
Objectspace.  They are probably the only piece of software I am
impressed with because they're free and they look so much like
Smalltalk.  
I loved the Toplink product until we got screwed on the licensing.  It
would be nice if Cincom could aquire the Toplink for Smalltalk from
them before they screw it up.  
The real problem occurs when you try to integrate all this stuff.
Most EJB servers implement thier own orb that is not complient with
anything other than java, so it pukes when you throw in a third party
orb.  If you do get it to work you break something of theirs.  
Final word.  Don't let the licensing scheme of Cincom scare you off.
Like everyone they will work with you to acheive something that makes
both parties happy.  The Smalltalk world is small and they probably
know it they don't work something out with you, you will walk away
from the technology and they will never have that chance again.  In
the Java world I can get some slack because I am with a large company,
but I get the feeling sometimes they just as soon tell me to bugger
off!  

On a final note, Integration is always overlooked,  why do you think
all those people do Microsoft stuff.  It all works together.
Sometimes it explains the price of pain.  Again Integration is always
overlooked and I think that is what Smalltalks(VisualWorks) strongest
point is.  Looking at the list above, just imagine degugging that
stuff.

Don Stacy

On 17 Apr 2001 00:46:18 +0100, [hidden email] (A SERFer) wrote:

+---- [hidden email] wrote (Mon, 16 Apr 2001 08:50:32 -0400):
 | We all seem to feel that Smalltalk is not a popular as it should be, and
 | that anyone that does an unbiased review would agree (smug smug smug).
 |
 | What *should* we do to sell Smalltalk?
 +----

A number of replies to your question!

Choosing only one is difficult.  I think a few of the other
issues are symptoms of licensing, so I'll pick that, for now. ;)

--
Gary Johnson     [hidden email]
Privacy on the net is still illegal.
<a href=http://www.squeak.org>Tired of selfish technology monopolies?</a>

dspublic-REMOVE THIS AND HYPENS- <[hidden email]> writes:

Smalltalk is not cheap for me.  I'm writing client-side apps which use
ODBC.  Price?  If I used Forte as the IDE, nothing at all.  (I actually
use VisualAge for Java, but that is not expensive).

The large scale is not where hearts and minds are won.  Languages
are often (mostly?) successful because they can be taken up and used
by the *small* developer and individuals, including the academic
market.  These people then take their enthusiasms and skills into
industry.  Smalltalk got HUGE publicity when it was available
on the very small scale, and cheaply, as Smalltalk/V.

Smalltalk needs a new, cross-platform Smalltalk/V.

>
>On a final note, Integration is always overlooked,  why do you think
>all those people do Microsoft stuff.  It all works together.
>Sometimes it explains the price of pain.  Again Integration is always
>overlooked and I think that is what Smalltalks(VisualWorks) strongest
>point is.  Looking at the list above, just imagine degugging that
>stuff.

We are moving away from Microsoft stuff, as the target for what you
have to integrate to is constantly changing!

A SERFer wrote:
$500.00 per year per developer, plus 5% of product revenue (based on the
revenue of the Smalltalk portion of the product - we are flexible there
as well).  The percentage shrinks if you are willing/able to pay more up
front on an annual basis

Small Developer models (limited to 100 or 200 deployed seats
respectively) - cost $1999.00 or $3999.00 per year, up to three
developers.



--
James A. Robertson
Product Manager (Smalltalk), Cincom
[hidden email]
<Talk Small and Carry a Big Class Library>